AI Paper Atlas

Research map / Architectures and efficiency / CNN design

GPCA: A Probabilistic Framework for Gaussian Process Embedded Channel Attention

Jiyang Xie, Zhanyu Ma, Dongliang Chang, Guoqiang Zhang, Jun Hai Guo

TPAMI 2021 · Architectures and efficiency › CNN design · approach: Compact CNN architectures · 33 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Gaussian process prior to channel attention (ablation)GPCA (ResNet50) · CIFAR-100top-1 accuracy: 72.36 -> 73.93Improved
add Gaussian process prior to channel attention (ablation)GPCA (ResNet18) · ImageNet-32x32top-1 accuracy: 47.37 -> 49.58Improved
add Gaussian process prior to channel attention (ablation)GPCA (VGG16) · miniImageNetaccuracy: 81.30 -> 82.78Improved
optimize kernel parameters instead of fixing them (ablation)GPCA (ResNet34) · miniImageNetaccuracy: 79.92 -> 80.14Improved
optimize kernel parameters instead of fixing them (ablation)GPCA (VGG16) · CIFAR-10top-1 accuracy: 94.50 -> 94.51No clear effect
optimize kernel parameters instead of fixing them (ablation)GPCA (ResNet18) · ImageNet-32x32top-1 accuracy: 49.34 -> 49.58Improved
add GPCA module to base networkResNet50 · CIFAR-100top-1 accuracy: 71.88 -> 73.93Improved
add GPCA module to base networkResNet34 · miniImageNetaccuracy: 78.08 -> 80.14Improved
add GPCA module to base networkResNet18 · ImageNettop-1 accuracy: 70.40 -> 71.61Improved
add GPCA module to base networkResNet50 · ImageNettop-1 accuracy: 75.20 -> 77.57Improved
replace SENet with GPCAResNet18 · ImageNet-32x32top-1 accuracy: 47.89 -> 49.58Improved
add GPCA to ADL for localizationADL (VGG16) · CUB-200-2011 WSOLtop-1 localization accuracy: 45.30 -> 54.97Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy